The data streaming in and out of organizations from electrical and mechanical sensors, RFID tags, smart meters, scanners, mobile communications, live social media, and more results in staggering volumes of information. When all these sources are networked to communicate with each other – without human intervention – the Internet of Things (IoT) is born.
The IoT market is estimated to include nearly 26 billion devices, with a "global economic value-add" of $1.9 trillion by 2020, according to Gartner, and IDC forecasts nearly $9 trillion in annual sales by 2020. By all accounts, IoT is a new type of industrial revolution.
But to derive useful knowledge from the tide of streaming source data – and participate in this new economy – you must have analytics.
In traditional analysis, data is stored and then analyzed. But with streaming data, analytics must occur in real time, as the data passes through. This allows you to identify and examine patterns of interest as the data is being created. The result is instant insight and immediate action.
So before the data is stored, in the cloud or in any high-performance repository, the event stream is automatically processed. And using analytics to decipher streaming data as close to the device as possible creates a new realm of knowledge for many industries.
Let's look at a few examples.
The Internet of Things in health care
In health care, analyzing IoT data can result in increased uptime for machines that treat cancer, which means that patients are treated when they are scheduled. If a treatment time is missed, it can be up to 40% less effective, so reducing service interruptions is critical. By monitoring hundreds of sensors, identifying issues early and proactively correcting them, service personnel armed with the necessary information and parts arrive together, said Todd DeSisto, at the 2014 Axeda Connexion Conference,
Elekta, a Swedish company that provides equipment and clinical management to help treat cancer and brain disorders, has cited a 30% reduction in site visits because of such monitoring, said Martin Gilday, another presenter at the Axeda event.
With rising global populations and corresponding increases in disease and health care costs, the remote patient monitoring market doubled from 2007 to 2011 and is projected to double again by 2016, according to a report from Kalorama Information.
And what if these scenarios went beyond monitoring device status or patient conditions – to predicting machine reliability in advance of parts beginning to malfunction? Servicing would then move from being proactive to being optimized across each supplier's landscape of devices. And foreseeing patient problems before they even experience symptoms could avert adverse events altogether.
Event streams that know more than just existing conditions, and which evaluate future scenarios using advanced analytics, are now within the realm of possibility.
How do you apply predictive capabilities to IoT data? High-performance analytics environments are designed to examine complex questions and produce models. These algorithms are then coded into the data streams, along with any data normalization and business rules to detect patterns associated with the defined future scenarios. So in addition to monitoring conditions and thresholds, you can use the data stream to assess likely future events.
The Internet of Things in manufacturing
The automobile industry is stepping up development detection systems for imminent collisions to determine when to take evasive action. Based on radar and other types of remote technologies, driving conditions are monitored to assess – and ultimately avoid – collisions. These collision avoidance systems assess the likelihood of a collision event and automatically prescribe mechanical changes to the vehicle if the driver doesn't respond – including deceleration and external lighting changes. Potential accident reduction from wide deployment could surpass $100 billion annually in savings, said an article in McKinsey Quarterly.
In fact, the "Industrial Internet,"—a term coined by GE which combines physical machinery, networked sensors and software – has extensive use and promise in manufacturing, including production optimization, product development and aftermarket servicing. GE predicts $1 trillion in opportunity per year by improving how assets and resources are used and how operations and maintenance are performed within industrial industries, according to a 2013 article in Forbes.